Plasma Nitriding of CA-6NM Steel: Effect of H2 + N2 Gas Mixtures in Nitride Layer Formation for Low N2 Contents at 500 °C 559

Plasma Nitriding of CA-6NM Steel: Effect of H2 + N2 Gas Mixtures

in Nitride Layer Formation for Low N2 Contents at 500 °C

Angela Nardelli Allensteina, Carlos Maurício Lepienskib, Augusto José de Almeida Buschinellic, Silvio Francisco Brunattod,*

aEngineering Post-graduate Program (PIPE), Federal University of Parana – UFPR, CEP 81531-990, Curitiba, PR, Brazil bDepartment of Physics, Federal University of Parana – UFPR, CEP 80230-901, Curitiba, PR, Brazil cDepartment of Mechanical Engineering, Federal University of Santa Catarina – UFSC, CEP 88040-900, Florianópolis, SC, Brazil dDepartment of Mechanical Engineering, Federal University of Parana – UFPR, CEP 81531-990, Curitiba, PR, Brazil

Received: October 1, 2010; Revised: December 13, 2010

This work aims to characterize the phases, thickness, hardness and hardness profiles of the nitride layers formed on the CA-6NM martensitic stainless steel which was plasma nitrided in gas mixtures containing different nitrogen amounts. Nitriding was performed at 500 °C temperature, and 532 Pa (4 Torr) pressure, for gas mixtures

of 5% N2 + 95% H2, 10% N2 + 90% H2, and 20% N2 + 80% H2, and 2 hours nitriding time. A 6 hours nitriding

time condition for gas mixture of 5% N2 + 95% H2 was also studied. Nitrided samples results were compared with non-nitrided condition. Thickness and microstructure of the nitrided layers were characterized by optical microscopy (OM), using Villela and Nital etchants, and the phases were identified by X-ray diffraction. Hardness profiles and hardness measured on surface steel were determined using Vickers hardness and nanoindentation tester, respectively. It was verified that nitrided layer produced in CA-6NM martensitc stainless steel is constituted of compound layer, being that formation of the diffusion zone was not observed for the studied conditions. The higher the nitrogen amounts in gas mixture the higher is the thickness of the nitrided layer and the probability

to form different nitride phases, in the case γ’-Fe4N, ε-Fe2-3N and CrN phases. Intrinsic hardness of the nitrided layers produced in the CA-6NM stainless steel is about 12-14 GPa (~1200-1400 HV).

Keywords: plasma nitriding, Low N2 content gas mixtures, ASTM CA-6NM martensitic stainless steel, nanoindentation technique

1. Introduction The martensitic wrought stainless steels, CA-6NM, were of interaction with nitrogen. Elements as Al and Ti have a strong developed in Suisse to improve the weldability of conventional interaction with nitrogen, while the interaction of Cr is dependent martensitic stainless steels as, for example, the ASTM CA-15. of the alloying element content on steel. It is considered chromium Classified by ASTM as a soft stainless steel, CA-6NM was evaluate contents higher than 5.6 wt. (%) leads to strong nitrogen interaction to attend the market which needs steels with superior mechanical characteristics. 10 properties and facilities to fabricate. This stainless steel after Pinedo and Monteiro studied the plasma nitriding kinetics of tempering presents excellent mechanical properties, as high strength, AISI 420 martensitic stainless steel. Nitriding was performed at 480, cavitation erosion and adequate toughness, even at low temperatures. 500, 520, 540 and 560 °C temperatures, to 4 hours, using a gas mixture of 75% N + 25% H , and pressure of 250 Pa (1.87 Torr). For all the In addition, due to its high hardenability, CA-6NM is utilized in 2 2 nitriding temperatures, the nitrided surface was comprised by both the diverse applications, normally in the manufacturing of large pieces compound layer and the diffusion zone, even at lower temperatures, as hydraulic turbine rotors, pumps and compressors1-3. as a result of the high nitrogen potential on the gas mixture. The Otherwise, nitriding is a thermochemical surface treatment widely maximum nitriding surface hardness was 1500 HV0.025. The hardness used to treat steels and alloys. The main advantages of nitriding are profiles show a step shaped decrease from the maximum hardness the improvement of wear and frictional properties, corrosion and to the core hardness. This behavior is a consequence of the complex fatigue resistance4. Plasma nitriding5-7 is one of the most versatile nitriding reactions that take place at the interface, and reflects the nitriding processes with many advantages over the conventional salt- nitrogen compositional profile across the nitrided case. bath and gas nitriding, and it has been widely used in many industrial In the present work, surface hardness of the plasma nitrided applications to improve the wear and fatigue resistance. CA-6NM martensitic stainless steel samples was determined by As pointed out by literature8,9 the presence of alloying elements, nanoindentation technique, for different nitriding conditions, particularly nitride formers, has a strong influence on the hardening, considering different N2 + H2 gas mixtures and nitriding times, and morphology and kinetics characteristics of the nitrided surface. The results were compared to the obtained hardness profiles and they were influence of nitride former elements is dependent on the degree correlated to the phases present in the nitride layers.

*e-mail: [email protected] 558 Allenstein et al. Materials Research

2. Materials and Methods in solid solution in the substrate bulk, and the stainless characteristic is kept unaltered in such region. Otherwise, in the nitrided layer, the Samples were prepared from a cast CA-6NM martensitic stainless nitrogen diffusion into surface, allied to the high chromium-nitrogen steel, which was received in the 22 HRC hardness tempered condition. affinity at the considered temperature tends to result in chromium The chemical composition in wt. (%) was 0.032% C, 12.25% Cr, nitride precipitation. Such assumption is in accordance with the 4.42% Ni, 0.63% Mn, 0.522% Si, 0.43% Mo and Fe balance, and X-ray diffraction results presented ahead. As a consequence of the microstructure is comprised of martensitic matrix ( ) with some α the strong solid solution chromium depletion close to surface, the dispersed austenite ( . Before nitriding, the samples were cut for γ) nitrided layer was also revealed using Nital etchant. It is to be noted 20 × 30 × 10 mm dimensions, ground and polished using 1 m µ Nital is a non‑adequate etchant for stainless steels. Such result must Al O solution, in accordance with the conventional metallographic 2 3 be emphasized, since the use of Nital solution, in this case, would preparation. permit to indicate if the surface was nitrogen sensitized due to the The plasma nitriding apparatus is shown in Figure 1. Nitriding nitriding treatment for the considered material. Moreover, such result was performed at 500 °C temperature, at 532 Pa (4 Torr) pressure, indicates Nital solution as etchant could be utilized as a direct method gas flow of 240 sccm (4.00 × 10–6 standard m3/s), for 2 and 6 hours to determine if the stainless characteristic of the stainless steel in the nitriding times, using different gas mixtures of 20% N + 80% H , 2 2 nitrided layer region is maintained unaltered after nitriding treatments. 10% N + 90% H and 5% N + 95% H . The discharge chamber 2 2 2 2 The above mentioned is based confronting both the nitrided layer and consisted of a stainless steel cylinder 350 mm in diameter and 380 mm in height. Samples were placed in the cathode and were negatively bulk regions aspects at micrographs shown in Figure 2a2, b2, c2, d2. biased in a selected potential, using a square waveform pulsed In addition, the use of Nital solution facilitated the determination of the nitriding depth in the present work. Otherwise, Vilella etchant was power supply of 3.6 kW. The potential was specified to 660 ± 20 V and the power transferred to the plasma was adjusted by varying responsible by etching similarly the nitrided layer and bulk regions, the time switched on (ton) of the pulse. The ton of the pulse could resulting in visual contrast slight differences of both the regions, as be varied from 10 to 220 µs and the total on/off time was 240 µs. shown in Figure 2a1, b1, c1, d1. The sample temperature was varied by adjusting the on/off time It is to be emphasized nitriding layer is normally composed of of the pulsed potential and was measured using a chromel/alumel compound layer and diffusion zone. For the processing conditions (type K) thermocouple that is placed into the sample holder, which is and characterization techniques adopted in the present work it was inserted 12 mm into the sample. The 1.5 mm diameter thermocouple is electrically isolated with Al2O3 and is protected with a stainless steel cover. Prior to nitriding, the system was evacuated to a residual pressure of 2.6 Pa (2 × 10–2 Torr). The (99.999% pure) and (99.999% N2 H2 pure) gas mixtures were obtained adjusting datametrics mass flow controllers whose full scale value was the same for both, in such case 500 sccm (8.33 × 10–6 standard m3/s). The total gas flow was set to 240 sccm (4 × 10–6 standard m3/s) in order to maintain the specified atmosphere of the discharge. The pressure in the vacuum chamber was adjusted by manual valves and measured using a capacitance manometer of 1.33 × 104 Pa (100 Torr) full scale. Heating rates on the order of 15 °C/min was employed to reach the plasma nitriding temperature. Prior to nitriding, the samples were cleaned in a H2 glow discharge at a pressure of 400 Pa (3 Torr) and at a temperature of 300 °C during 10 minutes. The nitrided samples were cross-section cut for microstructural analysis. The surfaces were then ground and polished with 1 µm diamond paste, and after etched by two different etchants, Vilella and 2% Nital solution. Optical microscopy analysis was performed in a Olympus PME equipment. Microhardness measurements were performed using a HMV Microhardness Tester Schimadzu equipment, with 100 g load for profiles determination and 500 g load for surface measurements. The nanoindentation technique was executed on nitrided and non-nitrided samples. The equipment utilized was a Nanoindenter XP - MTS Systems, with 40 g (400 mN) load and 10 seconds loads time. Finally, X-ray diffractometry (XRD) analysis was performed on surfaces of nitrided samples, using a Shimadzu equipment, for 1.54 nm wave-length (λ) copper tube. 3. Results and Discussion Figure 2 shows typical microstructures after the nitriding treatments, which were etched by Vilella (Index 1) and Nital (Index 2). It is possible to verify the Nital etching made possible to reveal the nitrided layers for all the studied conditions. The same was not verified for the respective substrate bulks. This result clearly indicates as the nitriding process develops, the chromium alloying element is kept Figure 1. Plasma nitriding apparatus representation. 2010; 13(4) Plasma Nitriding of CA-6NM Steel: Effect of H2 + N2 Gas Mixtures in Nitride Layer Formation for Low N2 Contents at 500 °C 559

Figure 2. Micrographies for different nitriding conditions: a) 20% N2 – 2 hours; b) 10% N2 – 2 hours; c) 5% N2 – 2 hours; and d) 5% N2 – 6 hours (Index 1: etched with Vilella; and Index 2: etched with Nital 2%). 560 Allenstein et al. Materials Research

not possible to observe any diffusion zone. Such fact could be related increase of nitriding time for 6 hours, using 5% N2 gas mixture, to its very small thickness, due to short nitriding times utilized here. resulted in hardness of 1173 HV0.5 (Table 1). This assumption is in agreement with Li and Bell results11, which Figures 3 and 4 show hardness profiles of processed samples as a demonstrated a very thin diffusion zone was obtained in martensitic function of the gas mixture and nitriding time, respectively. Hardness stainless steel for 20 hours nitriding time at 500 °C (10 and 3.33 times values of 800, 800 and 500 HV0.1 were measured at depth of 10 µm, higher, and at the same nitriding temperature utilized here). inside the compound layer, for 20, 10 and 5% N2 gas mixtures and Table 1 shows the thickness of the nitriding layer obtained for 2 hours processed conditions, respectively. The increase of nitriding all the studied conditions. It can be verified the higher the nitrogen time for 6 hours, using 5% N2 gas mixture, resulted in hardness content in gas mixture the higher is the thickness of the nitriding of 1100 HV0.1 at depth of 10 µm, inside the compound layer. For layer, the same occurring for higher nitriding times. Thicknesses of all studied conditions, measurements performed at 30 µm depth, comprising the hardness of the bulk, indicated values on the order 25.0, 20.0 and 12.5 µm were measured for 20, 10 and 5% N2 gas mixtures and 2 hours processing conditions, respectively. The increase of 300 HV0.1. Figure 5 presents result of nanoindentation as a function of the of nitriding time for 6 hours, using 5% N2 gas mixture, resulted in thickness of 25.0 µm. Table 1 also presents hardness measurements indenter penetration, which was obtained by varying the applied loads up to 40 g (400 mN). For non-nitrided condition, the high of the samples surface. Hardness values of 1240, 1172 and 950 HV0.5 were determined for nitrided layers produced after nitriding times hardness values about 4 - 5 GPa (~ 400 - 500 HV), related to indenter of 2 hours and using gas mixtures containing 20, 10 and 5% N , penetrations up to 500 nm, could be probably explained by the 2 occurrence of surface work hardening generated during polishing respectively. The compound layer produced using 5% N2 showed lesser hardness than 1200-1300 HV probably due to the elastic-plastic step, at the sample preparation procedure, which would be in 14 deformation field which reaches non-nitrided substrate (the bulk), as accordance with Pethica et al. study . For higher loads resulting in 2250 to 2500 nm indenter depth, the measurements values tend to the a consequence of the high testing load that was utilized, in the case substrate bulk hardness, in the case 3 GPa (~ 300 HV). For 20, 10, and 500 g, and the small layer thickness, which was only of 12.5 µm, as 5% N , and 2 hours nitriding conditions, the values of 13 (~ 1300 HV), previously presented. In such case, hardness tests must be carried 2 14 (~ 1400 HV) and 12 GPa (~ 1200 HV), respectively, are in out applying low loads to prevent this kind of error in measurement. agreement with the expected for surface hardness measurements. It is to be noted different studies have recommended that maximum Confronting such results and the obtained by Vickers hardness tester penetration depths do not reach 1/10 and 1/20 of the layer thickness, (Table 1), nanoindentation resulted in higher measurement values to respectively, for very hard and soft materials12,13. Otherwise, the the last two nitriding conditions (10 and 5% N2). This fact is related to the smaller thickness of the respective compound layers and the strong differences in the applied loads in both the characterization techniques, Table 1. Compound layer thickness and the respective surface hardness values in the case 40 g maximum load and 500 g for nanoindentation and for different nitriding conditions. Vickers hardness measurement techniques, respectively. In such Nitriding Thickness of Surface hardness case, the very high applied load utilized for hardness determination

condition compound layer (HV0.5) results in a higher penetration of the Vickers indenter, which probably (µm) overtook the compound layer, entering in the substrate bulk region. Finally, for 5% N and 6 hours nitriding condition, values of 12 GPa 20% N2 – 2 hours 25.0 1240 2 10% N – 2 hours 20.0 1172 (~1200 HV) obtained for penetrations higher than 500 nm are in 2 agreement with microhardness measurements (Table 1). The small 5% N – 2 hours 12.5 950 2 hardness values for penetrations up to 500 nm are probably due to 5% N – 6 hours 25.0 1173 2 the surface roughness, as a consequence of the longest time that the Non-nitrided - 250 sample was exposed to plasma species bombardment.

Figure 3. Hardness profiles of processed samples as a function of the gas Figure 4. Hardness profiles of processed samples as a function of the nitriding mixture. time. 2010; 13(4) Plasma Nitriding of CA-6NM Steel: Effect of H2 + N2 Gas Mixtures in Nitride Layer Formation for Low N2 Contents at 500 °C 561

Figure 5. Nanoindentation results for different nitriding conditions: a) Non-nitrided; b) 20% N2 – 2 hours; c) 10% N2 – 2 hours; d) 5% N2 – 2 hours; and e) 5% N2 – 6 hours.

The effect for which smaller hardness values are verified for small Figure 6 presents a comparative of the X-ray diffraction patterns penetrations in the nitrided layer was also verified for all nitriding obtained for different nitriding conditions. In the start material, conditions comprising 2 hours nitriding time (Figure 5b, c, d). The two identified phases are present, in the case α−(FeCr phase, variation of the hardness values along the first 300 nm is higher corresponding to a cubic martensite, which is a typical tempering as the nitrogen content in gas mixture is increased, which results product) and γ−(FeCrNi phase, which is normally present in the in stronger sputtering effect, leading to the obtainment of a higher CA-6NM stainless steel, for relative amounts up to 5%). All nitriding roughness surface. conditions resulted in formation of chromium nitride (CrN phase) 562 Allenstein et al. Materials Research

• The hardness of the nitrided layers formed on CA-6NM stainless steel surface ranged from 12 to 14 GPa at penetration depths about 500 nm; • The nanoindentation hardness results are in accordance with Vickers hardness measurements for the studied nitrided layers. • The hardness values as a function of the indenter penetration, for penetrations smaller than 500 nm, are strongly dependent of the studied processing conditions; and • Nital solution as etchant could be utilized as a direct method to determine if the stainless characteristic of the stainless steel in the nitrided layer region is maintained unaltered after nitriding treatments.

Acknowledgments The authors would like to thank the Brazilian Agency CNPq for financial support to perform this research, and LACTEC-UFPR by supplying the CA-6NM stainless steel samples.

References 1. Crawford JD. CA6NM an Update. In 29th Annual Steel Founder’s Society of America Technical and Operating Conference Proceedings. USA; 1974. p. 1-13. Figure 6. X-ray diffraction patterns for plasma nitrided CA-6NM stainless 2. Gooch TG. Heat Treatment of Welding 13%Cr-4%Ni Martensitic Stainless steel surfaces in different conditions. Steel for Sour Service. Welding Research Supplement. 1995; 213-222. 3. Bilmes PD, Llorente CL and Pérez IJ. Toughness and microstructure of and in the start phases peak enlargement. The peak enlargement is 13Cr4NiMo high strength steel welds. Journal of Materials Engineering more evident for α−(FeCr) phase, and the obtainment of expanded and Performance. 2000; 9:609-615. martensite has been considered in the present case. In addition, it is 4. Sun Y and Bell T. Plasma surface engineering of low alloy steel. Materials Science and Enginering A. 1991; 140:419-434. to be noted the 20% N2 and 2 hours nitriding condition also resulted in the formation of iron nitrides, which were identified as γ’-Fe4N 5. Edenhofer B. Physical and Metallurgical Aspects of Ionnitriding. Heat Treatment Metals Part 1. 1974; 23-28. and ε-Fe2-3N phases. The fact the iron nitrides precipitation only occurred for the 6. Rie K-T. Recent Advances in plasma diffusion processes. Surface and highest %N2 gas mixture must be explained. Considering the iron Coatings Technology. 1999; 112:56-62. nitrides are metastable phases, results indicate the probability of 7. Berg M, Budtz-Jørgensen CV, Reitz H, Schweitz KO, Chevallier J,

γ’‑Fe4N and ε-Fe2-3N formation is very small for nitriding conditions Kringhøj P et al. On Plasma Nitriding of Steels. Surface and Coatings comprising lower %N2 gas mixtures. In such case, smaller amounts Technology. 2000; 124:25-31. of nitriding reactive species could be available in glow discharge for 8. Jack DH and Jack KH. Invited review: Carbides and Nitrides in Steel. 5 and 10% N2 gas mixtures. An assumption the nitriding reactive Materials Science and Engineering. 1973; 11(1):1-27. species as atomic nitrogen would preferentially react with chromium 9. Lightfoot J and Jack DH. Kinetics of Nitriding With and Without alloying element, in the compound layer, is supported by the fact the Compound Layer Formation. In Proceedings of Heat Treatment’ 73. chromium forms stable nitrides, which present negative formation London, UK; 1973. p. 59-65. Gibbs Free Energy Variation (∆G) for the specified nitriding 10. Pinedo CE and Monteiro WA. On the kinetics of plasma nitriding temperature. So, the higher the nitrogen amounts in gas mixture a martensitic stainless steel type AISI 420. Surface and Coatings the higher is the probability to form different nitrides in compound Technology. 2004; 179:119-123. layer. It is to be noted the same argument, id est., a higher amount of 11. Li CX and Bell T. Corrosion properties of plasma nitrided AISI 410 available nitriding reactive species for higher %N gas mixtures can 2 martensitic stainless steel in 3.5% NaCl and 1% HCl aqueous solutions. be utilized to explain the nitrided layer thickness evolution as function Corrosion Science. 2006; 48:2036-2049. of gas mixture as previously presented (Table 1). 12. Johnson KL. The correlation of indentation experiments. Journal of 4. Conclusions Mechanics and Physics of Solids. 1970; 18:115-126. 13. Zielinski W, Huang H and Gerberich WW. Microscopy and The main conclusions of the present work can be listed as follows: microindentation mechanics of single crystal Fe-3 wt.% Si: Part II. • The higher the nitrogen amounts in gas mixture the higher is TEM of the indentation plastic zone. Journal of Materials Research, the probability to form different nitride phases in compound 1993; 8(6):1300-1310.

layer, in the case γ’-Fe4N and ε-Fe2-3N and CrN phases, and 14. Pethica JB, Hutchings R and Oliver WC. Hardness measurement at the higher is the thickness of the nitrided layer, for the studied penetration depths as small as 20 nm. Philosophical Magazine A. 1983; conditions; 48(4):593-606.